Resource exhaustion in Vllm
CVE-2026-57173
vLLM is an inference and serving engine for large language models. Prior to 0.24.0, the input_audio handling path for /v1/chat/completions calls AudioMediaIO.load_bytes or AudioMediaIO.load_file without passing VLLM_MAX_AUDIO_DECODE_DURATION_S to the shared audio decoder. An unauthenticated client can therefore submit a small compressed audio input that expands into a very large float32 PCM allocation, bypassing the duration guard already used by /v1/audio/transcriptions and causing an out-of-memory worker crash. Inline data URLs reach this path without being bounded by VLLM_AUDIO_FETCH_TIMEOUT. The issue affects deployments serving an audio-capable model, and authentication changes only the deployment-specific reachability. This issue is fixed in version 0.24.0.
Published · last modified .
CVSS v3 metric
CVSS v3 base score 6.5 (Medium). Vector: CVSS:3.1/AV:N/AC:L/PR:L/UI:N/S:U/C:N/I:N/A:H.
EPSS exploit prediction
EPSS: 0.007 (54.2th percentile), scored .
Very low probability of exploitation in the next 30 days; routine patching cadence is appropriate. 54th percentile — 54.2% of CVEs in the catalogue have a lower EPSS than this one. How to read EPSS.
EPSS over last 30 days · oldest: 0.007 · newest: 0.007 · change: +0.000
Affected products
- Vllm
- Vllm-Project Vllm — versions < 0.24.0
Weakness classification (CWE)
References
- github.com/vllm-project/vllm/security/advisories/GHSA-hcwq-8wjf-3gcr (Vendor Advisory)
- github.com/vllm-project/vllm/pull/45908 (Patch, Issue Tracking)
- github.com/vllm-project/vllm/commit/3d20275bb4d434f53055c3c0b645fd8bb072965e (Patch)
- github.com/vllm-project/vllm/releases/tag/v0.24.0 (Release Notes)
Frequently asked questions
- What is CVE-2026-57173?
- CVE-2026-57173 is a medium-severity vulnerability in Vllm, classified under Allocation of Resources Without Limits or Throttling. CVSS score: 6.5/10. Published 2026-09-16.
- How severe is CVE-2026-57173?
- Medium severity. CVSS v3 base score is 6.5 out of 10.